def _get_conv_layers(self, in_channel, conv):
for stage_num, stage_param in enumerate(self.__conv_num_list):
stage_name = "stage%d_" % stage_num
filter_size = stage_param[0]
layer_count = stage_param[1]
for layer_num in xrange(layer_count):
out_channel = stage_param[layer_num + 2]
with tf.variable_scope(stage_name + "conv%d" % layer_num):
weights = tf.get_variable("weights",
initializer=tf.truncated_normal([filter_size, filter_size,
in_channel, out_channel],
stddev=math.sqrt(2.0 / (filter_size ** 2 * in_channel))))
biases = tf.get_variable("biases",
initializer=tf.zeros([out_channel]))
conv = tf.nn.conv2d(conv, weights,
strides=[1, 1, 1, 1], padding="SAME")
conv = tf.nn.bias_add(conv, biases)
with tf.variable_scope(stage_name + "bn%d" % layer_num):
bn = batch_normalize(conv, convnet=True)
with tf.variable_scope(stage_name + "relu%d" % layer_num):
relu = tf.nn.relu(bn)
in_channel = out_channel
conv = relu
with tf.variable_scope(stage_name + "pool"):
pool = tf.nn.max_pool(conv, ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1], padding="SAME")
conv = pool
return conv
def conv2d(tensor_in,
n_filters,
filter_shape,
strides=None,
padding='SAME',
bias=True,
activation=None,
batch_norm=False):
"""Creates 2D convolutional subgraph with bank of filters.
Uses tf.nn.conv2d under the hood.
Creates a filter bank:
[filter_shape[0], filter_shape[1], tensor_in[3], n_filters]
and applies it to the input tensor.
Args:
tensor_in: input Tensor, 4D shape:
[batch, in_height, in_width, in_depth].
n_filters: number of filters in the bank.
filter_shape: Shape of filters, a list of ints, 1-D of length 2.
strides: A list of ints, 1-D of length 4. The stride of the sliding
window for each dimension of input.
padding: A string: 'SAME' or 'VALID'. The type of padding algorthim to use.
See the [comment here]
(https://www.tensorflow.org/api_docs/python/nn.html#convolution)
bias: Boolean, if to add bias.
activation: Activation Op, optional. If provided applied on the output.
batch_norm: Whether to apply batch normalization.
Returns:
A Tensor with resulting convolution.
"""
with vs.variable_scope('convolution'):
if strides is None:
strides = [1, 1, 1, 1]
input_shape = tensor_in.get_shape()
filter_shape = list(filter_shape) + [input_shape[3], n_filters]
filters = vs.get_variable('filters', filter_shape, dtypes.float32)
output = nn.conv2d(tensor_in, filters, strides, padding)
if bias:
bias_var = vs.get_variable('bias', [1, 1, 1, n_filters],
dtypes.float32)
output += bias_var
if batch_norm:
output = batch_normalize(output, convnet=True)
if activation:
output = activation(output)
return output
def conv2d(tensor_in,
n_filters,
filter_shape,
strides=None,
padding='SAME',
bias=True,
activation=None,
batch_norm=False):
"""Creates 2D convolutional subgraph with bank of filters.
Uses tf.nn.conv2d under the hood.
Creates a filter bank:
[filter_shape[0], filter_shape[1], tensor_in[3], n_filters]
and applies it to the input tensor.
Args:
tensor_in: input Tensor, 4D shape:
[batch, in_height, in_width, in_depth].
n_filters: number of filters in the bank.
filter_shape: Shape of filters, a list of ints, 1-D of length 2.
strides: A list of ints, 1-D of length 4. The stride of the sliding
window for each dimension of input.
padding: A string: 'SAME' or 'VALID'. The type of padding algorthim to use.
See the [comment here]
(https://www.tensorflow.org/api_docs/python/nn.html#convolution)
bias: Boolean, if to add bias.
activation: Activation Op, optional. If provided applied on the output.
batch_norm: Whether to apply batch normalization.
Returns:
A Tensor with resulting convolution.
"""
with vs.variable_scope('convolution'):
if strides is None:
strides = [1, 1, 1, 1]
input_shape = tensor_in.get_shape()
filter_shape = list(filter_shape) + [input_shape[3], n_filters]
filters = vs.get_variable('filters', filter_shape, dtypes.float32)
output = nn.conv2d(tensor_in, filters, strides, padding)
if bias:
bias_var = vs.get_variable('bias', [1, 1, 1, n_filters], dtypes.float32)
output += bias_var
if batch_norm:
output = batch_normalize(output, convnet=True)
if activation:
output = activation(output)
return output